There's one thing that nobody seems to have touched on here. The practice of bypassing power supply 'lytics with film caps originated with the audiophiles, their rationale being that 'lytics have some amount of series inductance, which can cause their impedance to rise slightly at higher frequencies. All caps have some amount of series inductance, but the audiophiles consider that of 'lytics to be high enough to be objectionable.
Of course, if you follow this line of reasoning fully, then you must also be making a parallel resonant circuit by shunting the inductance in the 'lytic with the film cap, and a parallel resonant circuit has a very high impedance at its resonant frequency, and you could end up with a power supply that "rings" at some high frequency.
But how much series inductance does a 'lytic really have? I don't know offhand. But if it's high enough to be objectionable at audio frequencies, then it's also high enough that bypassing with a film cap could cause it to be resonant in or near the audio band. Aye, there's the rub!
Here's some good reading about 'lytic characteristics:
http://www.chemi-con.com/u7002/characteristics.php
Hmm, they're saying that really big 'lytics can have series inductance as high as 34 NANOhenries. That's .034uH. About as much inductance as you'd get from a couple of turns of wire.
I would answer this question for myself with careful measurements (with test equipment), since all you're going to get on the Internet is opinions and conjecture (see above! :wink: ).